Identification and antibiotic susceptibility of bacterial isolates from probiotic products

Abstract

In the present study, a total of 55 European probiotic products were evaluated with regard to the identity and the antibiotic resistance of the bacterial isolates recovered from these products. Bacterial isolation from 30 dried food supplements and 25 dairy products, yielded a total of 268 bacterial isolates selected from several selective media. Counts of food supplements showed bacterial recovery in 19 (63%) of the dried food supplements ranging from 10³ to 10⁶ CFU/g, whereas all dairy products yielded growth in the range of 10⁵–10⁹ CFU/ml. After identification of the isolates using whole-cell protein profiling, mislabeling was noted in 47% of the food supplements and 40% of the dairy products. In six food supplements, Enterococcus faecium was isolated whereas only two of those products claim this species on their label. Using the disc diffusion method, antibiotic resistance among 187 isolates was detected against kanamycin (79% of the isolates), vancomycin (65%), tetracycline (26%), penicillinG (23%), erythromycin (16%) and chloramphenicol (11%). Overall, 68.4% of the isolates showed resistance against multiple antibiotics including intrinsic resistances. Initially, 38% of the isolated enterococci was classified as vancomycin resistant using the disc diffusion method, whereas additional broth dilution and PCR assays clearly showed that all E. faecium isolates were in fact vancomycin susceptible.

Introduction

The past 5 years have witnessed a strong expansion of the probiotic market and, in parallel, a rise in the number of research projects addressing fundamental and applied aspects of probiotics. New research technologies have supported earlier suggestions of health promoting properties of probiotic lactic acid bacteria (LAB) as reviewed by Naidu et al. (1999) including stabilisation of the intestinal microflora by competition against pathogens (Gibson et al., 1997), reduction of lactose intolerance (de Vrese et al., 2001), prevention of antibiotic-induced diarrhea (Pochapin, 2000), prevention of colon cancer (Wollowski et al., 2001), and stimulation of the immune system (Isolauri et al., 2001). Bringing a probiotic to the market involves a step-wise process that needs to be carefully monitored in order to obtain a correctly labeled, functional, and safe product Sanders and Huis in't Veld, 1999, Saarela et al., 2000. If a product is not labeled correctly, safety and functionality cannot be guaranteed due to lack of documentation of the product components. However, as many manufacturers rely on the widely acknowledged but occasionally debated GRAS (‘generally regarded as safe’) status of lactobacilli and bifidobacteria (Salminen et al., 1998), characterization of probiotic LAB strains with regard to taxonomic status, antibiotic resistance, and virulence may sometimes be neglected.

Microbial analyses of probiotic dairy products have demonstrated that the identity and the number of recovered species do not always correspond to the information stated on the product label Reuter, 1997, Holzapfel et al., 1998, Hamilton-Miller et al., 1999. However, it should be noted, that each of the cited studies was rather limited in number and type of products or was mainly restricted to national products. Various opinions exist as to whether it might be desirable that some probiotic strains show resistance to specific antibiotics that are, for instance, involved in antibiotic-induced diarrhea (Charteris et al., 1998). On the other hand, the commercial introduction of probiotics containing antibiotic resistant strains may also have negative consequences, for example, when resistance is transferred to intestinal pathogens (Curragh and Collins, 1992).

In the current paper, an extensive study is presented to verify the label correctness of a range of European probiotic food supplements and dairy products, together with the antibiotic susceptibility testing of the product isolates. For each of these products, the label information was checked through taxonomic characterisation of the recoverable bacterial strains using whole-cell protein profiling. In addition, individual susceptibilities were determined for a selection of six antimicrobial agents.